Predicting metal and metal mixture effects in aquatic biota.

Dixon, D.G.¹, U. Borgmann², W.P. Norwood^1,2, M. Nowierski¹ and J. Shroeder^1,3.

¹ University of Waterloo, Waterloo, ON;
² Environment Canada, Burlington, ON;
³ Ministry of the Environment, Toronto, ON.

Total metal concentrations in the environment do not provide accurate estimates of toxicological effects since toxicity is a function of metal speciation and bioavailability. The main objectives of this project are to determine 1) the best methodology for accurately predicting single metal effects and 2) the most appropriate method of quantifying the effects of metal mixtures. These objectives are to fit into an Ecological Risk Assessment (ERA) framework such that the resulting methods can be used in hazard identification, exposure assessment, effects assessment and finally risk characterization.

The project has been divided into three studies which examine the various interactions affecting metal solubility, availability, bioaccumulation and toxicity. The first study, through the development and testing of metal mixture models, examines the impact of metal-metal interactions on bioaccumulation and toxicity. Within this study individual exposure:bioaccumulation:toxicity relationship models and Critical Body Concentrations (CBC's) have been determined and/or reported for 10 metals. The individual metal models can be combined to formulate mixture models (such as Concentration Addition or Effects Addition models) and are used to determine any interactions between metals by identifying differential effects (bioaccumulation and/or toxicity). The second study, through the development of a Biotic Ligand Model (BLM), examines the interaction between water chemistry and bioaccumulation and will be linked with CBC's in order to predict effects (interaction between bioaccumulation and toxicity). The third study (sediment assessment), the field component of the project, examines the effect of overlay water on uptake of metals from field collected sediments through investigation of interactions between water chemistry, solubility and bioaccumulation. CBC's derived in the first study are used to identify metals of concern in the third study, and bioaccumulation observed in the different water types in the third study will be compared with the BLM model produced in the second study.

Integration of the final results of the three studies will be instrumental in providing tools for the Ecological Risk Assessment process. Hazard identification/problem formulation can be achieved from exposure assessment of site, water and sediment chemistry from which bioaccumulation and, ultimately, risk characterization can be predicted utilizing the BLM and metal mixture models. These predictions will be verified with the sediment toxicity test data collected during study 3.

Top of Page

The Biotic Ligand Model: Making it work in the real world

Wood, C.M., L. M. Taylor, C. Kamunde, C. Ho, B. Baldisserotto, S. Niyogi and D.G. McDonald

Department of Biology, McMaster University, 1280 Main Street West, Hamilton, Ontario, L8S 4K1

The Biotic Ligand Model (BLM) is an elegant and cost-effective framework which provides site-specific water quality regulations for cationic metals. The BLM approach is based on an understanding of the key toxic mechanism(s) of action of metals at the fills, and relates the predicted gill burden in a given water chemistry to predicted toxic effects. Present versions of the BLM are designed to meet the immediate needs of the US EPA to protect against acute waterborne toxicity, and calibration data area provided from just a few species (trout, fathead minnows, daphnia). However, in many jurisdictions, including those in Canada, environmental risk assessment and regulations are focused more strongly on preventing chronic toxicity to native species, where waterborne and dietary factors may play a role, and where the gill may or may not be the site of primary impact. The goal of our MITE-RN research at McMaster is to assess the importance of real world factors (e.g. dietary quality and quality, specific growth rates, water chemistry, tissue-specific metal burdens) for the development of both acute and chronic BLMS, as well as to develop the BLM approach for a key species, the yellow perch (Perca flavescens), endemic to metal-impacted waters of eastern Canada. The rainbow trout (Oncorhynchus mykiss) serves as the reference model. To date, our focus has been on Cd and Cu. This presentation will provide an overview of some of our ongoing projects, illustrating (i) a new in vitro method, for examining the fast compound of Cu-binding to gills, and the influence of water chemistry thereon; (ii) the influence of growth rate and ration on tissue-specific accumulation of waterborne Cd; (iii) the differential nature of relationships for gill Cd-binding versus toxicity and Ca versus Cd antagonism between yellow perch and rainbow trout; and (iv) the influence of dietary Na and Ca content on the uptake of Cu and Cd respectively at the gills, and the potential for fish to choose a diet which is protective against waterborne metal toxicity.

Top of Page

Links between tissue metal burdens in indigenous fish and metal induced effects at the organism and population levels

Campbell P.G.C., A. Giguère, L. Kraemer, Univ. du Québec, INRS-Eau; A. Hontela, A. Lacroix, A. Gravel, Université du Québec à Montréal, TOXEN; J. B. Rasmussen, G. Sherwood, J. Kovesces, A. Isles, Dept. Biology, McGill University.

This field project has been designed to test relationships between the physiological and population status of indigenous fish and (a) ecological factors (habitat quality, food resources), (b) toxicological factors (ambient and tissue metal concentrations) and (c) metal detoxification factors (metallothionein induction and subcellular metal partitioning). The key hypothesis tested in this project is that there exists a mechanistic link between the intracellular speciation of the metals and the manifestation of deleterious effects at the organism and population levels.

Metal partitioning. To minimize food-chain effects caused by the impoverished benthic communities present in the more contaminated lakes, juvenile YP (<10 g), a life stage that is exclusively planktivorous in the Rouyn-Noranda lakes, was sampled in the Rouyn-Noranda and Sudbury areas. A preferential accumulation of Cd and Cu in the subcellular liver fraction which includes metallothioneins, was shown. An increase in the contribution of mitochondria to total Cd burden suggested that mitochondria might be a site of toxic effects. We also observed a decrease in liver malondialdehyde with increasing tissue Cu. An enclosure experiment was carried out to determine the feasibility of using a habitat-swap experiment to study dynamics of metal uptake, metal elimination and MT induction in juvenile YP. Cages were set up in lakes Dufault (contaminated) and Opasatica (reference), and YP were either reciprocally transplanted or caged within their native lake as a control. Fish were sampled at various time points over 37 d; indigenous fish were also sampled. Liver, kidney, gills, digestive tract and carcass were collected for metal and metallothionein mRNA analysis (analyses in progress).
Physiology. Metal concentrations and physiological responses of YP (1+ and YOY, young of the year) were determined along the metal concentration gradient in Abitibi. The 1+ perch exhibited an exposure-dependent impairment of the cortisol stress response, tested in a confinement situation as well as following stimulation with ACTH in vivo, and an impaired capacity to mobilize liver glycogen. A gradient in body Cd was also detected in YOY but there was no evidence for impaired cortisol, T3 and T4 synthesis. There was no growth impairment in 1+ or YOY, in contrast to adult perch. In vitro experiments were completed to extend the field work, with 1+YP collected from a clean lake (Memphremagog, QC). The in vitro assay for testing the interrenal capacity to secrete cortisol has been adapted for 1+ YP and the sensitivity of the interrenal cells to Cd has been determined (LC50, EC50) and compared to adult YP, and rainbow trout. The experiments have been completed, the statistical analyses of the data are in progress.
Bioenergetics and benthic invertebrates. Size structure and metal analyses of the benthic samples collected in the Abitibi lakes (enumeration and identification) are in progress. The field work was extended to the Sudbury area in Year 3 to test the hypothesis that lakes along the recovery gradient from acidification and metal contamination will reflect the same gradation in bioenergetic impairment that we saw in relation to heavy metal exposure in the Abitibi region. Fourteen lakes were selected and sampled and a full range of size/age of YP were collected. Perch are being prepared for aging, stomach content analysis has begun, and LDH activity in muscle is being analyzed as an indicator of bioenergetic status.
Project C.3 will improve our understanding of the link between metal exposure and metal bioaccumulation, and between metal accumulation and the incidence of deleterious metal-induced effects. Together with project C.2 (Wood and McDonald), the results should lead to development and testing of the Biotic Ligand Model (BLM), a promising tool for predicting the links between exposure and accumulation and effects.

Top of Page

Food chain transfer of mercury in ecosystems impacted by mining/smelting activities, or by Hg from natural geological sources

Chan H.M¹., A. Scheuhammer¹, S. Weech², J. Elliott², and K. Cheng²

¹ MacDonald Campus of McGill University, Ste-Anne-de-Bellevue, QC
² University of British Columbia, Vancouver, BC

Some proportion of the inorganic mercury (Hg) originating from both natural and anthropogenic sources is subsequently methylated and accumulates in aquatic invertebrates, fish, and ultimately in piscivorous wild birds and mammals, and humans. The extent to which Hg from mining activities, or from natural geological sources, is methylated and taken up into aquatic food webs has not been well studied. Our research within the MITE-RN seeks to determine the risk of elevated methylHg exposure in fish-eating wildlife in areas impacted by mining/smelting activities, and also in some areas of Canada where there are high natural background concentrations of Hg from geological sources; and to study the biochemical interactions of Hg and Se in target tissues of animals exposed to methylHg. In the summer of 2001, continuing collections of small fish, water, and surface sediments were undertaken from 7 lakes in the Rouyn-Noranda area; 8 lakes in the Clyde Forks, Ontario area; and 5 lakes in the Pinchi, BC area. In addition, in the Pinchi area, blood and feather samples were collected from adult and nestling bald eagles from nests around Pinchi, Fraser, Stuart, Tezzeron, and Great Beaver Lakes. Hg levels in fish and eagles were compared with levels of Hg in environmental media, such as lake sediment and surrounding soils; and to lake water chemistry variables (eg - pH, DOC) known to influence Hg accumulation in fish. An important conclusion of this research is that the presence of elevated levels of inorganic Hg in the environment from typical geological sources is not, in itself, sufficient to cause elevated levels of Hg to accumulate in fish and fish-eating wildlife, especially when environmental chemistry does not favor Hg methylation (eg - high alkalinity, high pH conditions). With respect to Hg-Se interactions in tissues, good correlations were observed between Hg and Se concentrations in various tissues ofcommon loons, and bald eagles, two top level avian predators associated with aquatic food chains. Hg and Se interactions were also studied in experiments in which captive ring doves were dosed with different dietary levels of methylHg and Se. The significance of Se in the modulation the toxicity of Hg will be discussed.

Top of Page

Quantitative synthesis of aquatic environmental effects studies for metal mines of the Precambrian Shield

Grapentine, L., P. Jarvis

National Water Research Institute, Environment Canada, Burlington, ON

Determining the responses of natural communities of aquatic biota exposed to emissions of metals and metalloids from mining activities is an important component of an ecological risk assessment of metal production. While studies focussed on individual mining and smelting sites commonly address this objective, broader application of the information obtained can be limited by site-specific conditions and low statistical power. The technique of meta-analysis is a way of combining the data from separate studies to quantitatively evaluate the overall effects. Meta-anlysis is superior to traditional narrative or "vote-counting" reviews because it assesses the orignal data rather than the conclusions. Information from the individual studies is converted into a standardized measure of effects, allowing the data to be pooled and statistically analyzed. It is particularly suited to situations where the expected magnitude of an effect is small or variable across studies, conditions likely to apply to metal-exposed environments.

The purpose of this study is to determine if, on the balance, there are significant alterations of biological communities at sites exposed to mining activities and, if so, in what way and to what degree. The steps involved in the meta-analysis of metal mine biological effects studies will be described, including:

Assembly of the data;
Conversion of data within studies into common measures of community response to exposure to mining activities (i.e., effect sizes);
Combination of effect sizes from all studies to produce estimates of the overall effect size; and
Determination of the significance of the overall effect size.

A critical stage in the assembly of the data to to screen out studies that do not aquequately address to purpose of the meta-analysis.

Results of analyses completed on studies available to date will be presented.

Top of Page

Impacts of metal-contaminated forest soils from the Canadian shield on terrestrial organisms.

N.C. Feisthauer, G.L. Stephenson (ESG International, Guelph, ON), J.I. Princz (University of Guelph, Guelph, ON) and R.P. Scroggins (Environment Canada, Ottawa, ON)

Investigation of the effects of metal contamination in forest soils on terrestrial organisms focused upon two objectives. The first was to determine the level of toxicity of impacted forest soils from three sites on each of two established transects, to a battery of terrestrial tests using single-species sublethal toxicity and functional assays. The second was to assess the applicability of Environment Canada's terrestrial toxicity methods to forest soils and to determine if any methodological modifications were required. Soils were collected from locations along transects established downwind of smelter emissions in Rouyn-Noranda, Quebec, and in Sudbury, Ontario. Soils from Rouyn-Noranda (RN) were sampled in June 2001 while soils from Sudbury (SUD) were sampled in both June and October 2001. The concentrations of the metals of concern measured in the site soils demonstrated a gradient of Ni, Zn, Pb, Cu and Cd soil concentrations among the three sites within each transect (high contamination, low contamination and control or background sites). Tests with a plant, an earthworm, and an arthropod (collembola) species were conducted with undiluted site soils and the site soils with high metal contamination diluted with the corresponding control (background) site soil. No acute toxicity to earthworms or collembola was observed following screening tests with the undiluted RN soils. However, plant growth was significantly reduced in soils from the control site and the site with the high contamination. Adverse effects to all three species were observed from chronic exposure to soils from the control and high contamination sites relative to those from the site with the low contamination. The RN transect sites did not provide an acceptable toxicity gradient, and the relationship between pH, metal contamination, and soil type was ambiguous; therefore, definitive toxicity testing proceeded with Sudbury soils only. No acute toxicity to earthworms or collembola was observed following screening tests with the Sudbury soils, with the exception of toxicity to collembola in the control (background) soil collected in October in which naturally occurring fungi were abundant. Adverse effects to earthworm and collembola were observed in chronic tests with the Sudbury soils. The soils collected in October from the Sudbury control (background) and high contamination sites adversely affected the survival and reproduction of collembola.. Plant toxicity was observed following both acute and definitive (longer-term) exposure to the Sudbury site soils. The toxic effects observed corresponded to the metal contamination gradient. Environment Canada's test methods for assessing acute and sublethal responses in invertebrates and plants were applied successfully with forest soils. Only minor modifications of the test methods were necessary. The changes principally focused on soil preparation required prior to testing (i.e., drying and sieving of soils).

The data derived from this research should be used in a complimentary manner with data from MITE research being conducted under the Processes Domain. The toxicity data generated in this study should be looked on as a first step towards determining the relationship between metal levels in the soil environment and the potential risk to terrestrial organisms associated with the deposition of metals.

Top of Page